Graphene

Graphene is a single layer of graphite and the basic carbon modification of fullerenes and CNTs, as shown in Figure 2.1. These carbon structures can be formed by rolling up or folding a graphene sheet. The first description of graphene as carbon allotrope is from the 1940s and described the possible potential in their electrical/mechanical properties [45]. Due to the isolation of graphene monolayers by Geim and Novoselov [6] at the beginning of the 21^th

Scientific and technological background

6

even more. Due to the planar structure consisting of sp² carbon bonding, graphene exhibits high mechanical (1 TPa) [46], thermal >5000 W/mK [47]

and electrical properties with a calculated electrical resistivity of approximately 1.0 μΩcm [7].

Graphene offers a high application potential such as e.g. in polymer composites for increasing electrical/mechanical properties [48,49], as energy storage systems [50] such as supercapacitors [51], as fuel cells [52], as bio sensors [53] or in photovoltaic cells [54].

Figure 2.1: Graphene as fundamental structure for carbon allotropes [7].

Scientific and technological background Technologically, a distinction can be made between single-layer graphene (SLG), few-layer graphene (FLG), graphene nanosheets (GNs) and graphene nanoribbons (GNRs) [53]. An electrical resistance of approx. 3.35 μΩcm was calculated for FLG [55], whereby graphene exhibits individual properties as a function of the number of layers. Furthermore, graphene shows a temperature-dependent electrical behaviour, whereby the conductivity increases with increasing temperature, which is described as semiconductive behaviour [56].

The electrical conductivity of graphene decreases continuously with an increasing number of layers from 0.8∙10⁶ S/m (2 layers) to 0.18∙10⁶ S/m (9 layers) [55]. This behaviour can be also observed for the thermal conductivity. The change in the thermal conductivity with increasing number of layers is explained by the change in the phonon dispersion and thus in more resulting phonon states for “Umklapp skattering” [57,58]. An increased number of layers from 2 to 10 layers leads also to a reduction of the Young modulus from 980 GPa to 940 GPa, respectively. This is explained by the weaker van der Waals forces between the individual layers, which makes FLG unstable [59].

Growth model of graphene

Graphene can be synthesised by various manufacturing methods. These can be distinguished into two basic procedural principles, the chemically derived from graphite [60,61] which are followed by sonication and the synthesis in the CVD process [62] with metals as substrate. Similar to CNTs, the same catalysts are used to synthesise graphene (see Chapter 2.1.2), such as copper

Scientific and technological background

8

The synthesis of graphene takes place in several steps, as shown in Figure 2.2a,b. The different solubility of carbon into the substrates (nickel (a) and copper (b)) leads to a difference between the growth mechanism [64]. Wang et al. reported that the growth of graphene begins preferably at the grain boundaries of the substrate on its surface [65].

By using nickel as substrate (Figure 2.2a), first the adsorption of carbon precursors occurs (1), followed by the dehydrogenation on the substrate surface (2). In the next step carbon diffuse into and through to the substrate material (3). At energetically favorable points takes the diffusion to the substrate surface place (4). Finally, the segregation process of graphene starts.

Whereas, copper is used as substrate for the growth of graphene, as shown in Figure 2.2b, desorption of precursors occurs on the substrate surface (1). The segregation process of graphene takes place immediately, without any diffusion process (2) [64].

Scientific and technological background

Figure 2.2: Schematic growth mechanism of graphene in the CVD process on a nickel substrate (a) and on a copper substrate (b) [64].

Based on the chemical reactions of the precursor (methane - CH4) and the dehydrogenation of methane on the surface (s), with injected hydrogen (H2) according to equation 2.1, 2.2 and the chemisorption of hydrogen on the substrate surface [64], hydrogen has a decisive influence on the quality of graphene.

𝐶𝐻₄(𝑔) → 𝐶𝐻(𝑠) + 4𝐻(𝑠) (2.1) 𝐻₂(𝑔) → (𝑠) + 2 𝐻(𝑠) (2.2)

Scientific and technological background

10

The presence of hydrogen is important for the synthesis of graphene. It was shown that hydrogen reduces the crystallinity and the growth rate of graphene especially on copper [66]. In addition, other studies show that the synthesis of graphene cannot work without the presence of water, because it acts as an activator of the surface and as an etching reagent [67].

2.1.2 Carbon Nanotubes

Carbon nanotubes are another allotropy of carbon and can be described as a rolled-up graphene sheet. Carbon nanotubes can be distinguished between single wall (a-SWCNT), double wall (b-DWCNT) and multi wall (c-MWCNT) carbon nanotubes as shown in Figure 2.3. Since their discovery in the 1950s [2] and their description by Iijima approx. 40 years later [3], these structures have become increasingly more researched to identify application fields and to improve already known materials in their properties with CNTs.

Because of the C-C bonds and their σ orbitals in the plane direction, CNTs have a high potential in the formation of mechanical, thermal and electrical properties. Additionally, the remaining p-orbital forms a π-bond perpendicular to the CNT direction. By this sp² hybridization, the carbon atoms arrange in hexagonal lattices. However, this requires a zero-defect packing of the carbon atoms. In particular, the growth mechanism of CNTs has a considerable influence on the formation of physical properties by formation of defects such as bonding, hybridisation or lattice defects, such as sp³ orbitales [68]. Experimental studies have shown that CNTs, like graphene, also have high electrical conductivity [69] and high mechanical [70] properties. For defect-free CNTs a theoretically Young modulus of ~1 TPa is calculated [71].

Studies on the influence of defects in CNTs shown a decrease of the mechanical properties [72,73]. The strength and the failure strain can be

Scientific and technological background reduced by as much as 46 % and 80 % [72] and the tensile strength by 60 % [73]. Based on the presence of defects, the strength of CNTs can be reduced up to 36 GPa [74].

Figure 2.3: CNT modifications: SWCNT (a), DWCNT (b), MWCNT (c).

Scientific and technological background

12

The wall thickness, or rather the number of carbon layers has an influence of the properties. It could be shown that with increasing wall number the electrical conductivity of CNTs increases and a plateau is reached, which is confirmed by the decrease of the calculated reduction of the band gap energy [40]. The mechanical stiffness of CNTs was simulated and showed that with increasing wall number the Young modulus decreases [75]. Based on the measured specific resistance 𝜌𝑅 of CNTs films, the band gap energy Eg of CNTs can be calculated with a simple relation with the present temperature T and the Boltzmann constant kB as shown in equation (2.3). The band gap energy is reduced with increasing number of layers [39].

𝜌_𝑅= exp(𝐸_𝑔/2𝑘_𝐵∙ 𝑇) (2.3)

Growth model of carbon nanotubes

The growth of CNTs is observed via electron microscopic methods. Based on these observations the first growth models were created by Iijima in the early 1990s [19,20]. On the basis of these models and new analytical methods such as the ETEM [23], the growth of CNTs during their synthesis was observed more in detail.

Three methods have proven as particularly suitable for the production of CNTs. Especially the arc discharge [76] and laser vaporisation [77] is suitable for large scale production of CNTs [78]. In addition, the use of the CVD process [79] is also applied to synthesise CNTs. The growth of CNTs is divided into two different mechanisms. On the one hand the base-growth which also called as extrusion or root growth and the tip-growth mechanism [22,80]. For both mechanisms the presence of a catalyst required. As a catalyst for CNT growth, predominantly metals are used, eg. Fe, Ni, Co [81]

Scientific and technological background as well as ZnO [82], Au [83], Ni/SiO2 [25], which are introduced during the synthesis.

Both growth mechanisms in the CVD process are based on a physically fundamental mechanism that is described in Figure 2.4a. In the first step, the gaseous hydrocarbon molecules are adsorbed on the catalyst surface (1).

Whereupon the dissociation of the precursor (2) and its diffusion on or into the surface of the metal catalyst (3) is followed. Finally, the nucleation and the formation of the carbon structure (4) take place [84]. As described by Lo et al. base-growth (Figure 2.4b) or the tip-growth mechanism (Figure 2.4c) depends on the ratio of the substrate temperature TS to the gas temperature TG. If TS is higher than TG, the base-growth mechanism is dominant, whereas TG is higher than TS tip-growth mechanism follows [85]. The basics of the CVD process are described in chapter 2.3 in detail.

Scientific and technological background

14

Figure 2.4: Growth model for CNTs (a) [84], growth model for the base-growth mechanism (b) and base-growth model for the tip-base-growth mechanism (c) after [85].

Besides to experimental studies, analytical studies, using Computational Fluid Dynamics (CFD) simulation is an effective tool to determine temperature and flow behaviour in the CVD process for the production of CNTs and gained in importance of their description [86,87]. Based on these data, the process can be optimised to generate CNTs of a higher quality [87].

Scientific and technological background 2.1.3 Overview of Raman spectroscopy on carbon structures

For fundamental characterisations of carbon structures and their structure-property relations, Raman spectroscopy has emerged as an effective analysis method over the past decades. Raman spectroscopy provides the possibility of the effects of various treatment methods, such as doping [88,89], functionalisation [90–92], graphitisation of carbon structures [41–44] or the number of layers [40,93] in relation to the resulting properties. However, the penetration depth of laser using during Raman measurements is in the range of few nanometers [94].

Characteristic for sp² hybridized carbon structures is the so-called G-band, which is about 1550-1605 cm^-1 and the so called G'-band or 2D- band between 2500 and 2800 cm^-1 [95–98], as shown in Figure 2.5a.

The G-band stands for the in-plane oscillation of the carbon atoms at which they move towards each other and can be referred to as C-C stretching mode [96]. Ideally, the peak for pure carbon materials is at 1582 cm^-1 as described for graphite [98]. The existence of the G'-band indicates a high orientation of carbon atoms in hexagonal lattices and is based on electron-photon and electron-phonon interactions. Therefore this band is referred to as second order Raman scattering, whereas the formation of the G-band is only based on an interaction and can thus be regarded as first order Raman scattering [96,98].

The atomic order of sp²-hybridised carbon is disturbed by defects such as, lattice defects, which leads to a further band, as can be seen in the Raman spectra. This so-called D-band is located at ~1350 cm^-1 [95–98] (Figure 2.5a-damaged graphene). The intensity and position of the D-band as described by Behler et al. has in particular been dependent on the energy of the used laser

Scientific and technological background

16

to detect structural changes [98,99] which are determined by the above-mentioned treatment methods and can be referred to as defect densities. A pronounced formation of the D-band indicates the presence of sp³-hybridized carbon [100]. Especially in the case of SWCNTs, the formation of the radial breathing mode (RBM) occurs between 50 cm^-1 and 700 cm^-1 (Figure 2.5a).

This is an out-plane oscillation of the carbon atoms and occurs only in closed structures such as CNTs [96–98].

For metallic or semiconducting CNTs, a different formation of the G-band in the Raman spectrum occurs as shown in Figure 2.5b. The metallicity of CNTs results in a wide G-band formation compared to semiconducting CNTs. The thermal treatment of carbon structures such as thermal annealing leads to a structural change at atomic level, which affects the formation of Raman spectra. This is explained in detail more in 2.3.1.

Scientific and technological background

Figure 2.5: Raman spectra of different carbon structures (a) [95,98], Raman spectra of highly orientated pyrolytic graphite (HOPG) metallic and semiconductive SWCNTs (b) [95].

2.1.4 3D carbon foams -synthesis and applications-

3D carbon structures are characterised by their low densities coupled with high specific surfaces. Based on these facts they are called as carbon aerogels or carbon foams. Due to their 3-dimensional morphology, these structures have an extraordinarily high potential in the future compared to graphene or CNTs regarding to their properties and associated potential application fields.

These 3D carbon structures consist in their substructure of CNTs or graphene, which can be produced by various methods.

Scientific and technological background

18

Synthesis of 3D carbon foams

Since the description of carbon aerogels at the end of the 1980s by Pekala [101], carbon aerogels steadily gain in importance. The first aerogels were produced based on a sol-gel process with a subsequent freeze-drying [101].

According to Campbell [102], different synthesis methods for the manufacturing of 3D carbon aerogels can be distinguished in three synthesis methods, the direct synthesis, the template-free assembly of graphitic structures or a precursor conversion. Table 2.1 shows a summarised overview about the manufacturing methods for 3D carbon foams.

The direct synthesis of carbon aerogels is based on template or template-free manufacturing processes. However, the template-based methods can further split in a removal mechanism of the template via etching or reduction, such as in Aerographite [28]. The most favorable method to synthesise carbon aerogels is based on the removal of the template by etching after the carbon deposition and the reapplication of the template morphology, usually using nickel [27,103–109], metal oxide foams [110] or hybrid materials [111].

Carbon aerogels can also be produced template-free by interconnection of CNTs [29,112] or carbon nanofibers (CNFs) [113].

The assembly of graphitic structures is one of the most promising methods for the synthesis of carbon aerogels. This production method can be subdivided into a template-based [30,114–118], template-free [31,32,101,119–129] or substrate-based [130] synthesis. The methods have in common the use of graphene, graphene oxide and CNTs, which are first dissolved in a solution. In the case of template-based methods, these particles are deposited on the substrate surface and then dried [30,114–116].

Thereupon, the template is removed by an etching process or remains as a hybrid in the structure.

Scientific and technological background One of the most promising methods for the production of 3D carbon aerogels is the synthesis via sol-gel freeze drying [31,101,119–124]. Additionally, hybrids with metal oxides [125–127] or CNTs [128] can be used for manufacturing. These methods are based on the mixing used resorcinol and formaldehyde mixture and the gelation of this mixture and supercritical drying. In a last step this aerogel is pyrolysed in a tube furnace to get a carbon aerogel as reported by Pekala [101]. The structure of these aerogels is randomly oriented and consist of interconnected graphene sheets. Several groups developed a manufacturing method for 3D carbon aerogels based on a single step freeze drying process [32,129].

Table 2.1: Extract from the overview of different manufacturing process of 3D carbon foams after Campbell [101].

Direct

Scientific and technological background

Freeze drying Graphene oxide sponge [32,129]

On a

substrate

Graphene aerogel

assembled on Zn foil [129]

Especially the development of the synthesis of Aerographite [28,131] which is based on the replication of the template morphology in the CVD process,

Scientific and technological background shows an important impact on the research of 3D carbon foams. Therefore, the fundamentals of the CVD process are discussed in chapter 2.2. In this unique manufacturing method, ZnO is reduced and removed, so that only a hollow carbon shell remains. However, the occurred replication process in its details is still unknown. This replication process was modified by varying the carbon supply rate, which resulted in different densities and sub-morphologies [28].

These modifications in the morphology of Aerographite are resulting in different properties [35]. The density of Aerographite increases with increasing supply of carbon source. Thereby evolving the morphology from a hollow-framework over a closed-shell to closed-shell filled variant.

Resulting of the increased density, the compressive Young´s modulus and the electrical conductivity of Aerographite increases as well [35]. The phenomenon that with increasing density the mechanical properties increase at the same time is already described for foams by Ashby [132]. This linear correlation of the modulus and the electrical conductivity is related to the interconnection of single Aerographite tetrapods [35].

Applications for 3D carbon foams

Based on their high specific surface areas (SSAs) and low densities 3D carbon structures are particular suitable for supercapacitors [30,31,37,124,133], fuel cells [103], lithium Ion batteries [109], electromagnetic interference shielding [134] or catalysts [135–140].

Whereas, supercapacitors and catalysts are the most promising applications for 3D carbon structures and are therefore described in more detail below.

Scientific and technological background

22

Due to the storage mechanism supercapacitors, which are referred as electric double-layer capacitor (EDLC), are characterised by low charge and discharge times at high specific power densities of 15 kW/kg compared to conventional energy storage systems such as Li-ion accumulators (2 kW/kg) [140]. However, EDLCs have comparatively low specific energy densities.

EDLCs are suitable for applications in stationary or mobile systems such as car acceleration, emergency systems, and tramways. They are characterised by their long service life, which results from the number of charging and discharging cycles [141]. Electrochemical properties of carbon-based supercapacitors are summarised in Table 2.2. For carbon aerogels, capacities of up to 816 F/g [105] and SSA of up to 1873 m²/g [142] are achieved. The type of used electrolyte is a key parameter for the determined capacities. The capacity for an aqueous electrolyte at 45 F/g and for an organic electrolyte can be significantly lower (10 F/g) [124]. The capacity C of the 3D carbon structures is calculated via equation (2.9) [37,128].

C = ^{𝐼 ∙ ∆𝑡}

𝑚 ∙ ∆𝑉 (2.4)

With, the loaded current I in A, the discharge time Δt in s, the mass of active material m in mg and the potential change during the discharge process ΔV.

The capacity and performance of the EDLC dependence on the pore design and, on the specific surfaces area. Wang et al. showed that structures with the same pore properties but different designs (2D or 3D pores) influence the capacitive properties of the electrodes [143]. In addition, in another study, the linear dependence of the capacity on the achieved SSA was observed [144].

Scientific and technological background Table 2.2: Overview of properties of 3D carbon-based supercapacitors.

Material

Scientific and technological background importance of catalysts with a global demand worth US of $33.5 billion for the economy is summarised in a market study from 2014 [148].

In order to be able to use carbon aerogels for catalytic applications, catalytically active materials such as metals or metal oxides must be added to the carbon structure. For sol-gel-based carbon aerogels, a distinction can be drawn between 3 different manufacturing methods for catalytic aerogels after Moreno-Castilla et al. as shown in Table 2.3 [135].

The first method for the production of aerogels, catalytically active additives can be added to a mixture of formaldehyde and resorcinol, followed by a supercritical drying [136,137]. In the second method, the gels as described by Pekala [101] are produced and supplemented by catalytic additives. This is followed by a carbonization of the metal-loaded organic aerogels at high temperatures in a nitrogen atmosphere [139].

In the third method, catalytic metals are applied to the organic aerogel by adsorption, sublimation and supercritical deposition [135,138,140].

Scientific and technological background Table 2.3: Overview of synthesis methods for carbon foam-based catalysts.

Dissolving in initial mixture

Polymerisation of resorcinol derivative

Deposition of metal precursor SSA in m²/g 500-2240 /

1-176 689 -712 483-655 /

629-889 Catalysts Ce, Zr / Cr,

Fe, Co, Ni Cu Pt / Pt

References [136] / [137] [139] [138] / [140]

Scientific and technological background

26

2.2 Fundamentals of the chemical vapour deposition (CVD) -thermodynamic and mechanisms-

For the production of thin films different sub-types of the CVD process are available. In this case, the CVD method is used with regard to its temperature heating methods: Plasma enhanced CVD (PECVD) and thermal CVD, which can be distinguished in a hot-wall and a cold-wall CVD. Furthermore, the CVD process can be classified by the applied system pressures [149]. In the hot-wall CVD process, the reactor is heated by means of resistance heaters, so that the complete reactor chamber is heated. Here the deposition of the injected material takes place both on the template and on the reactor wall. In the case of the cold-wall CVD reactor, only the substrate is heated to the required temperature and the deposition of precursors occur just on the substrate [149].

The growth of thin film during thermal CVD processes are based on the nucleation of atoms from the vapour phase on a template surface [149–153].

The formation of the films is following fundamental key steps [149,151–

The formation of the films is following fundamental key steps [149,151–

Im Dokument Fundamentals of the growth mechanism, tailored properties and applications of 3D hollow carbon foams (Seite 23-0)